The Science Behind Spaced Repetition

The Forgetting Curve: Your Brain's Natural Decay

In 1885, German psychologist Hermann Ebbinghaus made a discovery that would change how we think about learning. He memorized lists of nonsense syllables and tested himself at various intervals. What he found was striking: memory decays in a predictable pattern.

Within 20 minutes, you forget about 40% of what you learned. Within a day, you've lost about 70%. Within a week, you're down to maybe 25% of the original memory.

This is the forgetting curve, and it's working against you every time you learn something new.

Fighting Back: The Spacing Effect

But Ebbinghaus discovered something else. Each time you review information just before you forget it, the memory becomes stronger. The forgetting curve flattens. The next review can wait longer.

This is the spacing effect, and it's the foundation of all spaced repetition systems.

Think of it like watering a plant. Water too often, and you waste effort. Water too rarely, and the plant dies. But water at just the right intervals, and the plant thrives with minimal effort.

From Theory to Practice: How SRS Works

A spaced repetition system (SRS) automates this process. It tracks when you learned each piece of information and predicts when you're about to forget it. Then it schedules a review at the optimal moment—not too early (wasted effort), not too late (forgotten).

The basic idea is simple:

• You review a card
• You rate how well you remembered it
• The system calculates when to show it again
• Easy cards get longer intervals
• Hard cards get shorter intervals

Over time, well-learned cards might appear once a month, while struggling cards appear every few days.

The Problem with Traditional Algorithms

Most SRS apps use algorithms designed decades ago. The original SuperMemo algorithm (SM-2) was created in 1987. It works, but it's based on assumptions that don't hold for everyone:

• It assumes everyone forgets at the same rate
• It uses fixed interval multipliers
• It doesn't learn from your actual performance
• It can't adapt to different types of material

If you're better at remembering kanji than vocabulary, SM-2 doesn't know. It treats everything the same.

FSRS: A Modern Approach

FSRS (Free Spaced Repetition Scheduler) is different. Developed by researchers who analyzed millions of real review records, it uses a mathematical model that actually learns how you forget.

Here's what makes it special:

Personalized forgetting curves - FSRS calculates your personal memory decay rate. Some people have "stickier" memories than others. FSRS adapts to you.

Difficulty estimation - Instead of fixed difficulty ratings, FSRS estimates how hard each card is based on your actual review history. A card you always struggle with gets recognized as difficult, even if you marked it "easy" once.

Optimal intervals - Using your personal parameters and each card's difficulty, FSRS calculates intervals that target a specific retention rate (usually 90%). This means you're always reviewing at the most efficient moment.

Continuous learning - The more you review, the better FSRS understands your memory. It keeps refining its predictions based on your real performance.

The Numbers Don't Lie

Studies comparing FSRS to traditional algorithms show significant improvements:

• 20-30% fewer reviews needed for the same retention
• More accurate predictions of when you'll forget
• Better handling of cards you find unusually easy or hard

For language learners doing hundreds of reviews daily, this efficiency gain is massive. It means more time for new content and less time grinding old cards.

Why This Matters for Japanese

Japanese has a unique challenge: you're learning multiple interconnected systems simultaneously. Kanji, vocabulary, grammar, reading, listening—they all reinforce each other, but they also have different forgetting patterns.

A smart algorithm like FSRS can recognize that:

• You remember kanji meanings better than readings
• Vocabulary from anime sticks better than textbook words
• Grammar patterns need more frequent review than individual words

This adaptability makes your study time dramatically more effective.

Making It Work for You

The best algorithm in the world won't help if you're not consistent. Here's how to get the most from spaced repetition:

1. Do your reviews daily - Skipping days creates a backlog that's hard to recover from. Even 10 minutes daily beats an hour once a week.

2. Be honest with ratings - Don't mark cards "easy" when you hesitated. The algorithm needs accurate data to help you.

3. Trust the system - If a card keeps coming back, it's because you need it. Don't get frustrated—get curious about why it's not sticking.

4. Add new cards steadily - A constant flow of new material keeps things interesting and builds your knowledge consistently.

The Bottom Line

Spaced repetition isn't magic—it's applied cognitive science. By reviewing at optimal intervals, you're working with your brain's natural patterns instead of against them.

And with modern algorithms like FSRS, you're getting the most efficient review schedule possible, personalized to how your specific brain works.

That's not just smart studying. That's studying smarter.